Effector T cell lines derived from rMT-II strain are highly sensitive to MBP stimulation in vitro

4. RESULTS

4.3 Effector T cell lines derived from rMT-II strain are highly sensitive to MBP stimulation in vitro

MBP-reactive CD4 effector T cell line can be established from immunized Lewis rats and are widely used for the induction of tEAE (Ben-Nun et al. 1981). Effector T cell lines can be readily established from rMT-II rats. Upon exposure to MBP-presenting APCs these cell lines vigorously proliferate and they are strongly pathogenic after transfer to healthy recipient rats (Lodygin, Kitz et al. unpublished). During generation and propagation of such lines, T cells are subjected to multiple rounds of antigenic stimulation and become terminally differentiated. We asked how the in vitro antigenic response of effector rMT-II T cell lines differs from the response of primary rMT-rMT-II T cells isolated ex vivo from naïve donors.

To measure the antigenic response, equal numbers of resting effector T cells and primary T cells isolated from blood or lymph nodes were co-cultured with thymic APCs loaded with MBP at different concentrations. Then we performed a time course analysis of cytokine production by ELISA and qPCR over 48 h time period (FIG 4A-C). At each tested concentration of antigen, rMT-II effector T cells rapidly (from 6 h after stimulation on) increased secretion of IFNγ and the concentration of this cytokine remained high until the end of experiment. Primary rMT-II LN T cells showed only transient and very small peak of IFNγ production around 12 h time point. A similar discrepancy between effector T cell line and primary LN T cells was observed for the IL-2 cytokine release (FIG 4B).

The transcriptional response to MBP stimulation was much higher in the effector cell line than in blood or LN derived T cells (FIG 4C). It reached its peak after around 6 h, and lasted until 24 h for IFNγ and 12 h post stimulation for IL-2 mRNA (FIG 4C). The

55 transcriptional response in blood and LNs T cells, even if much lower, was detected much later, at around 36 h after stimulation (FIG 4C).

The activation of LNs and blood rMT-II T cells by thymic APCs resulted in a very limited proliferation 72 h after stimulation, whereas these T cells vigorously proliferated in response to anti-CD3 and anti-CD28 antibody stimulation (FIG5 A). Additionally the CD25 expression in these cells increased after 48 hours and reached its peak in 72 hours with at all the antigen concentrations used. The maximal TCR stimulation by antibodies resulted, as expected, in a uniform increase of CD25 expression (FIG 5B).

FIG 4 In vitro activation kinetic of different rMT-II T cell subsets.

Time course of cytokine production determined by ELISA (A,B) and cytokine gene expression determined by quantitative reverse transcription polymerase chain reaction (RT-qPCR) (C). Concentration of IFNγ (A) and IL-2 (B) in supernatants collected at the indicated time after addition of antigen loaded APC was measured over 48 h period. MBP-specific effector line established from rMT-II (Teff cell line, in red) and primary T cells isolated from naïve rMT-II LNs (LN T cells, in green). The stimulation was performed in vitro with irradiated thymocytes in the presence of 100 µg/ml, 10 µg/ml and 1 µg/ml MBP. Data are mean

± SD. (C) RNA isolated from stimulated cells at the indicated time point after stimulation was converted to cDNA and analyzed by TaqMan PCR with primers and probes specific for IFNγ and IL-2 gene transcripts.

Shown is relative expression (mean ± SEM) normalized to the level of beta-actin gene expression. On the bottom plots the expression values for the cell line are not shown in order to reveal small scale changes in the gene expression detected in primary T cells. The number on the plot legend indicates concentration of antigen (µg/ml).

FIG 5 Activation of rMT-II T cells in vitro analyzed by flow cytometry.

(A) Proliferation-dependent dilution of fluorescent dye eFluor670 measured by flow cytometry. Primary rMT-II T cells isolated from blood and LN were stimulated with antigen loaded thymic APCs or with monoclonal antibodies specific for CD3 and CD28 (color code shown in panel B) and analyzed on day 3 after stimulation. (B) Expression of CD25 activation marker on blood and LN derived rMT-II T cells determined by immunostaining at the indicated time point after stimulation.

57 4.4 Phenotype and CNS residence of rMT-II T cells under steady state condition

The inability of rMT-II T cells to enter the tissue expressing the myelin target antigen in rMT-II rats might be a reason why these animals do not develop spontaneous autoimmunity. It is however known that the central nervous system, in spite being an immune-privileged organ, undergoes continues surveillance by immune cells (Deli et al.

2005). We therefore asked whether MBP-specific rMT-II T cells can be detected in the CNS of healthy rats. We analyzed T cell frequency in the spinal cord (SC) of healthy rMT-II rats (perfused with cold PBS to remove contaminating blood) and analyzed their immune phenotype. For comparison, we also analyzed T cells derived from immune organs such as blood, LN and non-immune organ, the gut.

In fact, we observed in the SC the presence of T cells. Expectedly, compared to blood, LNs or a preparation of intraepithelial lymphocytes (IELs) from the small intestine the numbers of these CNS-derived T cells were very low. However, we could counted an average of 1136.7 ± 245.6 T cells per g of the total SC tissue including the meninges, of which an average of 327.6 ± 59.2 T cells per g were present in the SC parenchyma (separated from meninges) (FIG 6A). This result indicates that T cells in rMT-II rats contribute, albeit at low rate, to CNS immune-surveillance. CD62L (L-selectin) expression can distinguish naïve T cells from memory T cells. Staining of T cells recovered from the SC revealed that the vast majority of them were CD62L negative (96.7 ± 1.5%) and therefore likely to be memory T cells (FIG 6B). In contrast, CD62L+ naïve rMT-II T cells were abundantly present in the blood and LN, where this population is known to circulate and home, respectively. Interestingly, among gut IELs 95% of T cells were also CD62L negative, which confirms the concept that T cells residing in peripheral organs have a memory phenotype (Gebhardt et al. 2013). Further, we found that CD25 expression was increased in T cells recovered from SC relative to the levels detected in blood and LN T cells (FIG 6C), consistent with the assumption that SC-derived T cells belong to the memory subset. This findings can be interpreted in two ways. First, memory T cells preferentially (if not exclusively) gain access to and maintain residency in the healthy CNS. Second, naïve rMT-II T cells enter CNS tissue, recognize locally presented antigen and differentiate into memory type, thereby ceasing expression of CD62L and upregulating CD25 expression. The second scenario is unlikely as the following data

58 suggest. As evident from TCR staining of rMT-II cells in different organs, there are two distinct populations of T cells with high and intermediate expression level of αβTCR (FIG 6A, see also FIG VII). TCR^int cells uniformly express the TCR-Vβ8.2 chain whereas TCR^high present with dispersed levels of Vβ8.2 expression and some of them express endogenous (non-transgenic) TCR chains at the levels equivalent to TCR expression in WT rats. Thus, TCR^high T cells in rMT-II rats most likely represent those cells that during development have escaped allelic exclusion and express a second TCR from the endogenous promoter that is somewhat stronger than the promoter driving transgene expression. Because MBP-specific effector T cell lines established from rMT-II rats become uniformly TCR^int (data not shown), T cells with high TCR expression seem to be less reactive to MBP and therefore are lost during repeated cycles of stimulation in vitro. Furthermore, when T cells isolated from naïve rMT-II rats are transferred into WT recipients, only the TCR^int fraction responds to the injection of cognate antigen by proliferation (FIG 6D), indicating that TCR^high cells react poorly to MBP. Since T cells recovered from the CNS are predominately TCR^high, it is unlikely that local recognition of MBP accounts for their CD62L- CD25+ phenotype.

FIG 6 Phenotype of rMT-II T cells in naïve animals.

Cells from blood, cervical lymph nodes (LNs), spinal cord (SC) and intraepithelial lymphocyte (IEL) preparation from the gut of rMT-II rats were isolated as described in METHODS, stained and analyzed by FACS to analyze T cell numbers and the expression of CD62L and CD25 surface markers.

(A) The frequency of αβTCR+ GFP+ T cells among gated living cells is representative of n > 10 samples in blood and LNs, n = 3 samples in the gut IELs and n = 6 samples in SC. The number of T cells infiltrating the SC and the SC parenchyma (SC par) is expressed as mean ± SD of n = 3 samples.

(B) The frequency of CD62L+ and - T cells is presented as mean ± SD of n = 5 samples in blood, LNs and SC and of n = 3 in gut IELs.

(C) The frequency of CD25+ T cells in blood and SC is expressed as mean ± SD of n = 3 samples. The histogram is representative of n = 6 experiments.

(D) The outline of experimental approach for in vivo testing antigen response of rMT-II T cells and representative flow cytometry results of measuring antigen-dependent proliferation in vivo (n = 6). CTV, Cell Tracer Violet; OVA, ovalbumin.

4.5 Intrathecal transfer of primary rMT-II T cells is not sufficient for the induction of clinical disease

Very low numbers of rMT-II T cells in the CNS and the prevalence of TCR^high fraction among them (presumably having no or very little MBP reactivity) may suggest that there are not enough self-reactive T cells entering the target organ to trigger the cascade of pathogenic events leading to EAE. To test this assumption, we decided to increase the number of T cells entering the CNS by performing an intrathecal injection of about 5 million rMT-II T cells isolated from a pool of LNs, blood and spleen into rMT-II rats. In another group of rMT-II rats we injected the same number of pathogenic effector T cell blasts of an rMT-II T cell line, harvested two days after MBP stimulation in vitro.

Interestingly, none of the rats injected with primary T cells developed EAE. In contrast, the rMT-II effector T cell blasts were able to induce a typical monophasic EAE course, starting on the third day after transfer, in all the injected animals (FIG 7A and B). In order to verify the presence of the transferred T cells in the CNS, rMT-II LNs-blood-spleen T cells were intrathecally injected in WT rats. As shown in FIG 7C, the transgenic green cells were found one day later both in the spinal cord and in the brain of these rats. Therefore, these results indicate that the functional state rather than the number of MBP-specific T cells crossing the BBB is crucial for the disease development.

FIG 7 Intrathecal injection of freshly activated effector but not primary rMT-II T cells induces EAE disease.

(A,B) rMT-II rats were injected intrathecally with 5x10⁶ rMT-II T cells. The clinical score and the percentage of weight variation were measured daily in each rat.

One group of rMT-II rats (n = 3) was injected with T cells of a rMT-II T cell line (in red). Another group was injected with rMT-II T cells magnetically isolated from a pool of LNs, spleen and blood (transferred rats n

= 6) of naïve rats (in green). (C) WT rats (n = 2) were injected intrathecally with 5 million rMT-II T cells isolated by MACS from a pool of LNs, spleen and blood of naïve rats. One day after transfer, the brain and the SC were extracted after perfusion, in order to quantify by FACS the GFP rMT-II T cells inside the central nervous system.

All data are presented as mean ± SD. In (C) is shown the mean ± SD (n = 2) of the GFP T cells number per total organ. Student’s t test was performed for all statistical analyses: In (A,B) is shown the statistical difference between the two groups in each highlighted time point.

62 4.6 Recently activated effector T cells show a much higher level of activation in the CNS than naïve or resting memory T cells

Our hypothesis that the functional state of the T cells dictates encephalitogenicity was further tested in a following set of experiments. We performed transfer or active EAE induction in rMT-II rats and analyzed the activation status of different T-cell subsets (FIG 8A). In the case of EAE induced by the transfer of TMBP-mCherry cells, we isolated T cells from the blood and inflamed CNS at the onset of disease. Under this condition we could distinguish three populations of T cells, namely effector T cells (Cherry⁺, GFP^-, CD62L^-), naïve rMT-II T cells (Cherry^-, GFP⁺, CD62L⁺) and memory rMT-II (Cherry^-, GFP⁺, CD62L^-).

In the case of active EAE induced by immunization, we also isolated T cells from the blood and inflamed CNS at the onset of disease. Under this condition we could distinguish two populations of T cells, these being effector/memory T cells (GFP⁺, CD62L^-) and naïve rMT-II T cells (GFP⁺, CD62L⁺) (FIG 8B).

In the transfer EAE experiments, effector TMBP cells were found in the blood and in the SC tissue at disease onset. As expected, the numbers of effector TMBP cells in this early clinical phase predominated over those of the recruited II T cells (FIG 8B). Recruited rMT-II T cells isolated from SC were almost exclusively CD62L negative, whereas in the blood about 80% were naïve (GFP⁺, CD62L⁺; FIG 8B). Similarly to transfer EAE, under condition of immunization nearly all rMT-II T cells entering the SC were CD62L negative (FIG 8B).

Therefore, naïve T cells are largely not capable of entering the CNS either in healthy condition or in the early stage of EAE disease.

In the transfer EAE experiment, the effector TMBP cells expressed higher levels of CD25 than rMT-II T cells both in the blood and in the SC (FIG 8C). The upregulation of CD25 expression in the SC relative to blood, which is indicative of antigenic stimulation, was stronger in effector TMBP cells than in rMT-II T cells (FIG 8C). However, in active EAE experiments, recently rMT-II T cells activated in the periphery also markedly upregulated CD25 upon entry into CNS (FIG 8D), very much like effector TMBP cells in the transfer EAE settings.

Further, we analyzed gene expression in T cells sorted by flow cytometry from blood and SC tissue in both types of EAE. As expected, effector TMBP cells isolated from SC tissue strongly upregulated the expression of pro-inflammatory cytokines IFNγ, IL-17 and GM-CSF (FIG 9). We could also confirm upregulation of mRNA of the IL-2 receptor alpha chain

63 gene (encoding CD25 protein, detected earlier by immunostaining). In contrast, rMT-II T cells recovered from SC tissue in transfer EAE showed only marginal increase of IFNγ, IL-17 and GM-CSF mRNA levels relative to levels in blood derived rMT-II T cells (FIG 9). In active EAE, rMT-II T cells in the SC tissue induced expression of pro-inflammatory cytokines much more strongly than did rMT-II T cells passively recruited in transfer EAE settings (FIG 9).

FIG 8 Transfer versus active EAE in rMT-II rats.

(A) In the first group EAE was induced in rMT-II rats by transfer of 4x10⁶ effector TMBP cells transduced in culture with retrovirus encoding mCherry red fluorescent protein (transfer EAE). The second group of rMT-II animals was immunized with MBP in CFA (Active EAE). When animals developed clinical EAE (day 3 for transfer EAE and day 6 for active EAE) spinal cord tissue and blood were analyzed by flow cytometry. In (B) is shown the frequency of αβTCR⁺ GFP⁺ (GFP T cells) and αβTCR⁺ mCherry⁺ GFP- (effector mCherry T cells) T cells in blood and SC after transfer EAE (top panels) and the number of GFP T cells in both the compartments in active EAE (bottom panels). Histogram shows the difference in CD62L expression of GFP T cells in the blood (blue) and in the SC, (red) both after transfer and after active EAE induction. In the overlay dot plots in (C) are compared the frequencies of CD25⁺ cells among the effector mCherry T cells (red) and the rMT-II GFP T cells (green) both in the blood and in the SC, after transfer EAE induction. The histograms in (C) show the difference in CD25 expression between the two groups of cells in the blood and in the SC, after transfer EAE induction. In (D) is shown the difference in CD25 expression between the rMT-II T cells in the blood and in the SC after active EAE induction. All the shown data are representative of n = 3 experiments.

FIG 9 Analysis of T cell activation in transfer and active EAE in TCR transgenic rMT-II rats.

Experimental setup is outlined in FIG 8A. The rats were sacrificed around day 3 after transfer, or day 6 after immunization. The cells were isolated from blood and spinal cord (SC), stained for αβTCR and sorted by FACS as αβTCR+ GFP+ T cells (in blue for Active EAE; in green for Transfer EAE) and as αβTCR+ mCherry+

GFP- T cells (in red). The graphs show the expression level of IFNγ, IL-17, CD25 and GM-CSF normalized to β-actin gene expression. All data are presented as mean ± SEM of n = 3 for active EAE and n = 3 and 6 for transfer EAE. The statistical analysis was performed by student’s t test. In the graphs are shown the statistical differences per each indicated pair of groups.

67 The entire population of rMT-II T cells in TCR transgenic rats is functionally heterogenic, i.e. it contains naïve and memory subsets and the memory subset itself may consist of central memory, effector memory and resident memory T cells. Moreover, the presence of TCR^high population with uncertain antigenic specificity complicates the interpretation of our results. Therefore, we decided to perform similar experiments in CD4 TMBP

“memory” rats, as this system allows an analysis of a more uniform population of T cells with well-defined antigenic specificity and differentiation history. To this end, we generated “memory” rats using GFP-tagged TMBP cells established in culture and neonatally transferred into WT Lewis recipients. These “memory” animals were used as a host for transfer of effector TMBP-Cherry cells to induce transfer EAE, or were subjected to immunization with MBP in CFA (FIG 10A). In transfer EAE settings, both recently activated effectors (Cherry⁺) and resting memory (GFP⁺) T cells were readily identified in blood at the day of onset of clinical symptoms, and both cell types entered SC tissue showing similar proportional enrichment (~5 fold) among all T cells detected (FIG 10B).

The CD25 expression of the two T cell subsets was similar in the blood and evidently increased in the SC in both T cell groups. Instead, the OX40 expression was different in the blood, but only the one in recently activated effector T cells increased in the SC (FIG 10B). Immunization of “memory” rats resulted in a prominent increase of GFP⁺ T cells frequency in the blood (from 0.7 to 5.9% of all T cells) and in the SC tissue (from 3.6 to 58.7% of all T cells) (FIG 10C). Furthermore, ex-memory TMBP cells recently activated in the peripheral lymph nodes by immunization, upon entry into SC tissue upregulated CD25 and OX40 surface activation markers relative to the levels in blood (FIG 10C).

FIG 10 Analysis of transfer and active EAE in TMBP-memory rats.

(A) Memory rats were prepared by i.p. injection of 2x10⁶ resting effector TMBP-GFP cells into newborn Lewis rats and 8 weeks later the engraftment was confirmed by flow cytometry analysis of peripheral blood samples. EAE was induced in the first group of “memory” rats by transfer of 4x10⁶TMBP-mCherry effector T cell blasts (transfer EAE) and in the second group by immunization with MBP in CFA (active EAE). The rats with clinical signs of disease were sacrificed 3 days after transfer (B) or 6 days after immunization (C). The cells were isolated from blood and spinal cord (SC) and analyzed by flow cytometry (B,C). (B) Shown is the frequency of GFP memory and mCherry effector T cells among the whole T cell population in blood and SC in transfer EAE. Numbers in the dot plots are frequencies of indicated T cell types (mean ± SEM of n = 5).

(C) Shown is the frequency of GFP memory T cells in active EAE. The dot plots are representative of n = 4.

Analysis of gene expression in T cells sorted from the CNS of memory rats at the onset of transfer or active EAE revealed substantial upregulation of IFNγ, IL-17 and CD25

Im Dokument The Role of Glycolysis in shaping the Autoimmune Potential of Myelin-Reactive T Cells in the Course of Experimental Autoimmune Encephalomyelitis (Seite 58-0)